Unpracticed Skills
If sight-reading is a learned skill, why can't I just learn it the way everyone else does?
One thing that Russell said in our interview went to the core of what I’m trying to do here:
This process of clumping is innate, of course, to any learned skill. But we forget that sight reading also requires that. You know when you learned the viola or violin, you had to do that. You went through such pain to figure out how you were going to coordinate the bow, and the left hand, and even to hold the instrument, let alone do all the billion things you do. But we forget that sight-reading also involves many different pieces that have to be learned.
This raises an obvious question: If sight-reading is a common skill among serious amateur chamber musicians (it is), and if it is a skill that involves many different pieces that have to be learned (as Russell asserts and as I believe), why is there not a well-developed road-map for learning to sight-read, in the way there’s a road map for learning how to play an instrument in the first place?
I’ve mentioned before that the prevalent advice in the amateur chamber music community is “if you sight read often enough, you’ll get the hang of it gradually”. Chamber musicians believe that sight-reading is learned through experience, but tend not to have a cultural awareness of how to improve at it through deliberate practice. By contrast, as Russell points out, everyone agrees that you shouldn’t learn to play the violin by playing through your pieces frequently and hoping to pick up technique through mere repetition— the techniques of practicing instrumental technique are among the foundational data sets for Ericsson’s account of deliberate practice.
As it turns out, Ericsson has also done a little work on deliberate practice in the area of sight-reading itself. Writing with Lehmann, he summarizes a situation that is similar to what Russell and I have each observed:
At times the ability to sight-read has been awarded a special status as indicator of musical skill, because its acquisition or origin is poorly understood. As Bean (1938) noted, the "successful learning of the skill of efficient reading seems to involve a trick of which neither teacher nor pupil is conscious" (p. 3). In addition, unlike the training of general performance skills, which is associated with many hours of structured training activities, good sight-readers do not report activities that are linked obviously to the acquisition of their superior skills. Despite the conspicuous absence of specific training activities, some trainability is usually acknowledged and musicians are encouraged to improve their skills for reading unfamiliar music (e.g., Bryant, 1997, p. 42; Hofmann, 1920/1976, p. 117; Roth, 1977, p. 41; Spillman, 1985,1990). However, some still contend that "teachers are merely guessing when they invent remedies for [enhancing sight-reading skills] that are based on no good sound theory" (Bean, 1938, p. 3).
…We investigate whether this specialized skill is deliberately acquired through engagement in relevant training activities or whether it is simply a function of talent or general level of instrumental proficiency.
I find it interesting that Lehmann and Ericsson think it’s necessary to demonstrate that sight-reading is developed through practice at all, rather than being a “function of talent or general level of instrumental proficiency”. I don’t know any musicians who think that experience sight-reading isn’t a necessary part of developing the skill. And as far as I know, everyone agrees that among expert musicians there’s a huge variation in sight-reading skill that’s independent of their general musical ability.1
In any case, Ericsson believes that all skills are developed through deliberate practice, so it is no surprise that he and Lehmann ultimately conclude that, indeed, sight-reading is not a function of talent or general level of instrumental proficiency, but rather:
[I]ndividual differences in sight-reading ability in our subjects and exceptional sight-reading feats by eminent musicians do not seem to reflect innate music talent or a specific sight-reading talent. Rather, they are the results of deliberate long-term involvement in relevant domain-related activities and appropriate self-imposed challenges.
In their study, they find that a significant amount of the variance in sight-reading ability among expert pianists can be explained by differences in experience accompanying and accumulated accompanying repertoire— hence their claim that sight-reading skill isn’t just a matter of innate talent but requires practice. They provide abundant quantitative evidence to support this uncontroversial view.
However, they go farther, and assert that experience sight-reading is not by itself enough to lead to expert sight-reading:
We reject the notion that improvement in performance is a direct consequence of mere sight-reading experience. For example, someone could play hymns in church for 20 years, and accumulate a great number of hours of accompanying experience and an enormous repertoire, yet all the repertoire is of the same complexity…
We believe that time spent in training activities may not be as important as the challenges that a pianist encounters during that time. Thus, according to the interview information provided by our subjects, the acquisition of sight-reading skills is associated with the performance of unfamiliar pieces, often under time constraint, starting with easier pieces early in their musical development and progressing to more difficult ones…
It is most likely in the process of engaging in challenging activities during sight-reading and in the process of rapidly acquiring repertoire that pianists adapt to the representative demands of the task. Here, they develop prerequisite skills such as the ability to represent and organize encountered music, thus facilitating subsequent recall or the ability to make informed guesses or improvise ("fake") probable sequences.
It seems to me that Lehmann and Ericsson are running into the same problem I am. They have clear quantitative evidence that experience with sight-reading leads to better sight-reading. When it comes to showing that mere experience is not enough, however, they allude vaguely to “interview information provided by our subjects” without providing any details, and otherwise provide an account that seems logical but speculative.
In any case, even according to their speculation, the process by which expert sight-readers acquire their expertise seems to be largely unconscious, and more in the realm of purposeful practice than deliberate practice as he has defined it. Ericsson identifies 3 general types of practice in Peak (I read the audiobook and can’t easily pull quotes, but here’s an excerpt from an online summary):
Not all practice is created equal. Doing something for a few hours isn't the same as practicing it deliberately.
There are a few different types of practice:
Basic practice. You keep doing something without thinking about it too much. The least useful type of practice.
Purposeful practice. You have clear goals and a plan in mind and practice outside your comfort zone. A much better way.
Deliberate practice. You learn from the best, with the best. Your practice routines are based on the best available knowledge and tailored specifically to you by your coach. You are laser-focused on building better mental representations and improving every little aspect of the skill. The most efficient practice method.
Using this vocabulary, most musicians seem to believe that sight-reading is acquired through basic practice, but Lehmann and Ericsson argue that good sight-readers are doing some purposeful practice as well, perhaps without thinking of it in those terms.
They reinforce my question and corroborate my impressions without providing clear answers. Their speculation does tend to reinforce my view that I should be trying to “develop prerequisite skills such as the ability to represent and organize encountered music”, presumably at the least by planning a sequence of escalating sight-reading challenges. I’d like to do this at the very least through a system of purposeful practice.
I’m working on reading some more about the psychology of sight-reading in various academic articles and possibly some books, but meanwhile I’ve been starting to read more about the psychology of reading text. I’m hoping to get a clearer picture of how sight-reading music might be similar to reading text, and what an explicit curriculum for teaching sight-reading might look like by analogy to the curricula that exist for teaching kids how to read English.
For instance, Lehmann and Ericsson again: “Biographers found it noteworthy to mention that outstanding pianists, such as Liszt, Czerny, and Mendelssohn, sight-read works in public early in their lives. However, not all outstanding pianists excel at sight-reading. There are several reported cases of expert pianists with deficient sight-reading skills (cf. Wolf, 1976, p. 143). Thus, there is considerable anecdotal evidence for the existence of large individual differences in the ability to sight-read among expert pianists.”